Problem Set #6 Power-flow solutions, Gauss-Seidel Method, and Newton-Raphson

Method

6-1 (Keyhani Lecture)

Using Gauss elimination and back substitution, solve
⎡10 3 0 ⎤ ⎡ x1 ⎤ ⎡1 ⎤
⎢ 4 12 2 ⎥ ⎢ x ⎥ = ⎢2⎥
⎢ ⎥⎢ 2 ⎥ ⎢ ⎥
⎢⎣ 5 2 14⎥⎦ ⎢⎣ x3 ⎥⎦ ⎢⎣3⎥⎦

Solution:
-Step 1
⎡10 3 0 ⎤ ⎡ x1 ⎤ ⎡ 1 ⎤
⎢ 0 10.8 2 ⎥ ⎢ x ⎥ = ⎢1.6 ⎥
⎢ ⎥⎢ 2 ⎥ ⎢ ⎥
⎢⎣ 0 0.5 14⎥⎦ ⎢⎣ x3 ⎥⎦ ⎢⎣2.5⎥⎦
-Step 2
⎡10 3 0 ⎤ ⎡ x1 ⎤ ⎡ 1 ⎤
⎢ 0 10.8 2 ⎥⎥ ⎢⎢ x 2 ⎥⎥ = ⎢⎢ 1.6 ⎥⎥
⎢
⎢⎣ 0 0 13.907 ⎥⎦ ⎢⎣ x3 ⎥⎦ ⎢⎣2.425⎥⎦
-Step 3
Back substitution:
2.425
x3 = = 0.1744
13.907
1.6 − 2 ⋅ 0.1744
x2 = = 0.1158
13.907
1 − 3 ⋅ 0.1158 − 0
x1 = = 0.06525
10

6-2 (Keyhani Lecture)

Use Newton-Raphson to find one solution to the polynomial equation f(x)=y, where y=0
and f(x)=3x3+4x2+5x+6. Start with x(0)=1.0 and continue until
x(i ) − x(i − 1)
<ε (6.1)
x(i − 1)
is satisfied, where ε=0.001.

Solution:
f ( x) = 3x 3 + 4 x 2 + 5 x + 6
df
J= = 9x 2 + 8x + 5
dx
[ ] [0 − f (x(i))]
x(i + 1) = x(i) + 9 x 2 (i ) + 8 x(i ) + 5
−1

Result:

1
x(0)=1, x(1)=0.18182, x(2)=-0.8636, x(3)=-1.432, x(4)=-1.284, x(5)=-1.265, x(6)=-1.265.
x(6) − x(5)
= 2 × 10 −4 < ε
x(5)
Stop after 6 iterations. Note that x=-1.26533.

The other two solutions are x=-0.034±j1.25677 (not required in this problem).

6-3 (Keyhani Lecture)

Use Newton-Raphson to find a solution to
⎡ e x1x2 ⎤ ⎡1.2 ⎤
⎢ ⎥=⎢ ⎥
⎣cos( x1 + x 2 )⎦ ⎣0.5⎦
where x1 and x2 are in radians.

a) Start with x1(0)=1.0 and x2(0)=0.5 and continue until Equation (6.1) is satisfied
with ε=0.005.
b) Show that Newton-Raphson diverges for the example if x1(0)=1.0 and x2(0)=2.0.

Solution:
a)
−1
⎡ x1 ⎤ ⎡ x1 ⎤ ⎡ x 2 e x1x2 x1e x1x2 ⎤ ⎡ 1.2 − e x1x2 ⎤
⎢x ⎥ =⎢ ⎥ +⎢ ⎥ ⎢ ⎥
⎣ 2 ⎦ new ⎣ x 2 ⎦ old ⎣− sin( x1 + x 2 ) − sin( x1 + x 2 )⎦ ⎣0.5 − cos( x1 + x 2 )⎦

x1 (0) = 1 , x 2 (0) = 0.5

i 0 1 2 3 4 5
x1 1.0 0.683 0.895 0.831 0.8267 0.8266
x2 0.5 0.386 0.151 0.215 0.2205 0.2206

x1 (5) − x1 (4)
= 1 × 10 − 4 < ε
x1 (4)
x 2 (5) − x 2 (4)
= 4 × 10 −4 < ε
x 2 (4)
b)
x1 (0) = 1 , x 2 (0) = 2

i 0 1 2
x1 1.0 6.655 -1.52×1028
x2 2.0 -10.147 5.21×1027
diverging

2
6-4 (Grainger and Stevenson, Jr. Chapter 8, Prob 8.10)

3
6-5 (Keyhani Lecture)
Consider the power system given below:

1) Assume the generator voltage is set at 10% above its rated value, and the load
connected to bus D is equal to 5 MW at a pf of 0.9 lagging. Compute the per-unit
model for the above system.
2) Compute the load voltage (Bus D).
3) Compute the active and reactive power supplied by Bus A.
4) Compute the active and reactive losses.

Solution:
1)
13.2 69
Sb=10 MVA, choose Vb , A = 132 = 13.2 kV , Vb , BC = 132 kV , Vb , D = 132 = 66 kV ,
132 138
132 2
Z b ,BC = = 1742.4 Ω
10

10 Z line 10 + j100
Z AB = j 0.10 = j 0.2 , Z BC = = = 0.006 + j 0.06
5 Z b , BC 1742.2
69 10
X T 2 = 0.08 ⋅ = 0.087 Z CD = jX T 2 = j 0.087 ,
66 10
5
S load = ∠ cos −1 0.9 = 5.56∠25.8° MVA = 5 MW + j2.42 MVar
0.9
S load , pu = 0.556∠25.8°

2) Z total = Z AB + Z BC + Z CD = j 0.2 + 0.006 + j 0.06 + j 0.087 = 0.006 + j 0.347

1
Y= = 2.88∠89° = 0.05 − j 2.8796
Z total

4
 ⎡Y Y12 ⎤ ⎡ Y −Y⎤
YBUS = ⎢ 11 ⎥=⎢
⎣Y21 Y22 ⎦ ⎣− Y Y ⎥⎦
⎡ 0.05 − j 2.8796 − 0.05 + j 2.8796⎤ ⎡ 2.88∠ − 89° − 2.88∠ − 89°⎤
=⎢ ⎥=⎢ ⎥
⎣− 0.05 + j 2.8796 0.05 − j 2.8796 ⎦ ⎣− 2.88∠ − 89° 2.88∠ − 89° ⎦

Gauss-Seidel Method:
S 2 = − S load , pu , which is the input complex power of Bus D
1 ⎡ S2 * ⎤
V2 (i + 1) = ⎢ − Y21V1 ⎥
Y22 ⎣V2 (i ) * ⎦
1 ⎡ − 0.556∠ − 25.8° ⎤
= ⎢ + 2.88∠ − 89° ⋅ 1.1∠0°⎥
2.88∠ − 89° ⎣ V2 (i ) * ⎦
- 0.193∠63.2°
= + 1.1
V2 (i ) *

Iteration number V2
0 1.0275∠-9.6544°
1 0.9998∠-8.6934°
2 1.0003∠-9.0418°
3 0.9993∠-9.0070°

V2 = 0.9993∠-9.0070°

3)
S1 = V1 (Y11 * V1 * +Y12 * V2 *)
= 1.1(3.168∠89° - 2.88∠89° ⋅ 0.9993∠ - 9.007°)
= 1.1(3.168∠89° - 2.878∠98°)
= 0.5017 + j 0.3492

Active power P1=0.5017×10=5.017MW

Reactive power Q1=0.3492×10=3.492MVar

4)
Active loss Ploss=P1-P2=5.017-5=0.017 MW
Reactive loss Qloss=Q1-Q2=3.492-2.42=1.072 MVar

5
Matlab code for the above calculation:

% EE740 Au04
% Problem 6-5 Gauss-Seidel
%
% 10/05/2004

V2(1)=1.0;
S2=-0.556*exp(25.8j*pi/180);
Y21=-2.88*exp(-89j*pi/180);
Y22=2.88*exp(-89j*pi/180);
V1=1.1;

for i=1:4
% V2(i+1)=(conj(S2)/conj(V2(i))-Y21*V1)/Y22;
V2(i+1)=-0.193*exp(63.2j*pi/180)/conj(V2(i))+1.1;
abs(V2(i+1))
angle(V2(i+1))*180/pi
end

Y11=Y22;
Y12=Y21;
S1=V1*(conj(Y11*V1)+conj(Y12*V2(5)))

6-6 (Keyhani Lecture)

Given:
Bus 1 is slack bus: V=1∠0°
Scheduled power at Bus 2 is 1.2 p.u.
Scheduled load at Bus 3 is 1.5 p.u.
Compute:
1) YBUS model
2) Bus voltages using Gauss-Seidel Method
3) Power mismatch at Bus 2 and 3
4) Power supplied by the swing bus
5) Power loss of transmission lines

6
Solution:
1)

⎡Y11 Y12 Y13 ⎤ ⎡ 14 − 4 − 10⎤

YBUS = ⎢⎢Y21 Y22 Y23 ⎥⎥ = ⎢⎢ − 4 9 − 5 ⎥⎥
⎢⎣Y31 Y32 Y33 ⎥⎦ ⎢⎣− 10 − 5 15 ⎥⎦

2) Gauss-Seidel Method
1 ⎡ Sk * k −1 N
⎤
Vk (i + 1) = ⎢ − ∑ YknVn (i + 1) − ∑ YknVn (i )⎥
Ykk ⎣Vk (i ) * n =1 n = k +1 ⎦

1⎡ S2 * ⎤
V2 (i + 1) = ⎢ − Y21V1 (i + 1) − Y23V3 (i )⎥
⎣V2 (i ) *
Y22 ⎦
1 ⎡ S3 * ⎤
V3 (i + 1) = ⎢ − Y31V1 (i + 1) − Y32V2 (i + 1)⎥
Y33 ⎣V3 (i ) * ⎦

i V2 V3
0 1 1
1 1.1333 0.9
2 1.0621 0.9333
3 1.0885 0.9224
4 1.0784 0.9180
5 1.0781 0.9171

V2=1.078 p.u., V3=0.917 p.u.

3)
S 2 = V2 (Y21 * V1 * +Y22 * V2 * +Y23 * V3 *)
= 1.078(-4 ⋅ 1 + 9 ⋅ 1.078 - 5 ⋅ 0.917)
= 1.2041

7
 S 3 = V3 (Y31 * V1 * +Y32 * V2 * +Y33 * V3 *)
= 0.917(-10 ⋅ 1 − 5 ⋅ 1.078 + 15 ⋅ 0.917)
= - 1.4993

Power mismatch at Bus 2

∆S 2 = S 2cal − S 2sch = 1.2041 − 1.2 = 0.041
∆P2 = P2cal − P2sch = 1.2041 − 1.2 = 0.041
∆Q2 = Q2cal − Q2sch = 0 − 0 = 0

Power mismatch at Bus 3

∆S 3 = S 3cal − S 3sch = −1.4993 − (−1.5) = 0.0007
∆P3 = P3cal − P3sch = −1.4993 − (−1.5) = 0.0007
∆Q3 = Q3cal − Q3sch = 0 − 0 = 0

4) Power supply by swing bus

S1 = V1 (Y11 * V1 * +Y12 * V2 * +Y13 * V3 *)
= 1 ⋅ (14 ⋅ 1 − 4 ⋅ 1.078 - 10 ⋅ 0.917)
= 0.518

For resistive power system, P1=S1=0.518, Q1=0.

5) Power loss of transmission lines:

S loss = Ploss = S1 + S 2 + S 3
= 0.518 + 1.2041 - 1.4993
= 0.223

6-7 (Grainger and Stevenson, Jr. Chapter 9, Prob 9.1)

8
9
6-8 (Keyhani Lecture)
Use Newton-Raphson Method to compute the bus voltages and power mismatches at Bus
1, 2, and 3 in Problem 6-6.

Solution:

⎡Y11 Y12 Y13 ⎤ ⎡ 14 − 4 − 10⎤

= ⎢⎢Y21 Y22
YBUS Y23 ⎥⎥ = ⎢⎢ − 4 9 − 5 ⎥⎥
⎢⎣Y31 Y32 Y33 ⎥⎦ ⎢⎣− 10 − 5 15 ⎥⎦
Assign initial condition of voltages: V2=1∠0°, V3=1∠0°

10
 ⎡ ∂P2 ∂P2 ∂P2 ∂P2 ⎤
⎢ ∂δ V2 V3
∂δ 3 ∂ V2 ∂ V3 ⎥
⎢ 2 ⎥
⎢ ∂P3 ∂P3
V2
∂P3
V3
∂P3 ⎥
⎢ ∂δ ∂δ 3 ∂ V2 ∂ V3 ⎥
J =⎢ 2 ⎥
⎢ ∂Q2 ∂Q2
V2
∂Q2
V3
∂Q2 ⎥
⎢ ∂δ 2 ∂δ 3 ∂ V2 ∂ V3 ⎥
⎢ ∂Q ∂Q3 ∂Q3 ∂Q3 ⎥
⎢ 3 V2 V3 ⎥
⎢⎣ ∂δ 2 ∂δ 3 ∂ V2 ∂ V3 ⎥⎦ 4×4
Because the transmission line model is only resistance and load complex power is also
the real power, therefore, δ2=δ3=Q1= Q2= Q3=0, then

⎡ ∂P2 ∂P2 ⎤
⎢ V2 ∂ V2
V3
∂ V3 ⎥
J =⎢ ⎥
⎢V ∂P3 ∂P3 ⎥
V3
⎢ 2 ∂ V2 ∂ V3 ⎥
⎣ ⎦ 2×2
So,
⎡ ∂P2 ∂P2 ⎤ ⎡ ∆ V2 ⎤
⎢ V2 ∂ V2
V3
∂ V3 ⎥ ⎢ V ⎥ ⎡ ∆P ⎤
⎢ ⎥⎢ 2 ⎥ = ⎢ 2 ⎥
⎢V ∂P3 ∂P3 ⎥ ⎢ ∆ V3 ⎥ ⎣ ∆P3 ⎦
V3
⎢ 2 ∂ V2 ∂ V3 ⎥⎢ V ⎥
⎣ ⎦ ⎣⎢ 3 ⎦⎥
Jacobian Correction Mismatch

The results of program (including Matlab source code) are shown below. The number of
iteration is 30.
⎡ 1.0 ⎤
V BUS = ⎢ 1.07749415 ⎥⎥
⎢
⎢⎣0.916750444⎥⎦

⎡ 0.02220446 × 10 −14 ⎤
⎢ ⎥
Pmismatch = ⎢− 0.02220446 × 10 −14 ⎥
⎢ 0 ⎥
⎣ ⎦

11
12
13